Comparison of the efficacy of intravenous tranexamic acid with and without topical administration versus placebo in urgent endoscopy rate for acute gastrointestinal bleeding: A double-blind randomized controlled trial

Abstract

Background

Tranexamic acid (TXA), a synthetic antifibrinolytic drug, is effective as a treatment for serious hemorrhage, including bleeding arising from major trauma and post-operative interventions. Significant acute gastrointestinal bleeding may have a poor outcome despite routine medical and endoscopic treatments. The aim of this study was to assess whether early intravenous and/or intravenous plus topical administration of TXA reduces the need for urgent endoscopy for acute gastrointestinal bleeding.

Method

This double-blind randomized clinical trial included 410 patients with proven acute gastrointestinal bleeding. All patients received conventional therapy. The subjects were randomized to three groups: (A) 138 patients received intravenous TXA (1 g q6h); (B) 133 patients received topical TXA (1 g single dose by nasogastric tube) plus systemic TXA; and (C) 139 patients received a placebo (sodium chloride 0.9%) for 24 hours. Subgroup statistical analyses were conducted for urgent endoscopy, mortality, re-bleeding, blood transfusion, endoscopic and/or surgical intervention rates, and health status.

Results

The time to endoscopy was significantly shorter in group C (15.58 ± 7.994, p < 0.001). A need for urgent endoscopy was seen in 14.49%, 10.52%, and 30.21% of patients in groups A, B, and C, respectively (p < 0.001). No significant statistical differences were seen between treatment groups regarding mortality, re-bleeding, blood transfusion, and endoscopic and/or surgical intervention rates. No thromboembolic event was documented during the 1-week follow up.

Conclusions

Our results showed that the antifibrinolytic properties of TXA can aid in changing an urgent endoscopy to an elective procedure, with better outcomes for both physicians and patients.

Keywords

Tranexamic acid gastrointestinal hemorrhage emergent endoscopy treatment

Introduction

Acute gastrointestinal bleeding (GIB) is a common life-threatening emergent condition that is associated with high morbidity and mortality rates.¹ Annually, acute upper GIB occurs in approximately 1 person per 1000, with a higher rate in males and the elderly.^2,3 About 65% of GIB originates from the upper gastrointestinal (GI) tract and the rest from the lower GI tract.⁴

Upper GIB management has progressed over past two decades, and yet GIB still has a high mortality.⁵ This point emphasizes the importance of differentiating patients who need urgent intervention (e.g. endoscopic, colonoscopic, surgical, or transfusion) from those who can afford to delay intervention or even be managed as outpatients. Endoscopy is currently regarded as an effective method for achieving therapeutic and diagnostic goals in the treatment of both upper and lower GIB. Pharmacological treatment also plays an important role, in addition to endoscopic hemostasis.⁶ The length of hospitalization, need for blood transfusion and intensive care unit (ICU) requirements can also be reduced by early endoscopy (defined as endoscopy performed within the first 24 hours of admission). Nevertheless, GIB clinical outcomes are not reduced by endoscopic intervention conducted within 6 hours of admission, and the intervention can also increase the length of hospital stays of patients who are clinically at high risk. Hence, rapid endoscopy should be done on a case-by-case basis.⁷

Tranexamic acid (TXA) has been recognized as an efficient pharmacological treatment with an antifibrinolytic effect that is useful for reducing hemorrhage following operations, bleeding trauma, or even heavy menstruation.^8–10 The results of the CRASH-2 trial showed that TXA administration significantly reduced both mortality rate (relative risk 0.91; 0.85–0.97) and death due to bleeding (relative risk 0.85; 0.76–0.96).¹¹ Accordingly, some studies also proposed the possible effectiveness of TXA as a treatment for acute GIB. Some previous trials have supported the use of TXA for reducing risk of death after GIB, while other trials showed no difference between patients who received TXA and those who received a placebo. One meta-analysis showed a reduction in mortality rate but no alleviation of bleeding or reduction in the need for transfusion.¹² A systematic review by Ker et al. reported that the topical use of TXA reduced bleeding and the need for blood transfusions in patients who underwent surgery.¹³ A recent study in a small sample of children with peptic ulcers reported that the use of topical TXA with an endoscopic procedure lowered the re-bleeding and the need for blood transfusions in the case group when compared with a control group.¹⁴ At present, the effectiveness of TXA in reducing the need for urgent endoscopy remains unestablished. However, the Hemorrhage Alleviation with Tranexamic Acid – Intestinal System (HALT-IT) trial is currently underway to determine the effect of TXA on the acute significant GIB mortality rate in about 8000 patients randomized for either TXA or a placebo.¹⁵

The aim of the present double-blind randomized trial was to assess the efficacy of intravenous TXA, with and without topical administration, versus placebo in reducing the need for urgent endoscopy, endoscopic and/or surgical intervention rates, blood transfusion requirements, and re-bleeding rates, and in improving the health at 72 hours and 1 month after admission in each group.

Methods

Study design

Patients older than 18 years with proven significant acute bleeding from any part of the gastrointestinal tract who primarily visited in the emergency room, ward and/or the ICU of Rasoul-e-Akram Hospital in Tehran, Iran were enrolled in this double-blind randomized controlled trial. Based on the study by Hawkey et al, the proportion of positive endoscopic findings was 52.7% and 26.3% in the patients who received placebo and TXA, respectively.¹⁶ In this regard, and considering type 1 error rate of 0.05 and power of 90%, the required sample size was estimated to be at least 135 patients for each group. The patients were randomly assigned to three groups using a computer-generated randomization table (138 patients in group (A), 133 in group (B), and 139 in group (C)). An independent researcher made random allocation cards using computer-generated random numbers. He kept the original random allocation sequences in an inaccessible third place and worked with a copy. Since the executors could get confused with the original coding of A and B later, the allocator recorded exactly what these codes meant to avoid further confusion. Because GIB is an emergency situation, brief information was collected on an entry form to evaluate patient eligibility. We needed to decide very quickly whether a patient was suitable for the trial, usually as soon as possible after the problem was identified. Group (A) received 1 g/q6h intravenous TXA, group (B) received a 1 g stat-dose by nasogastric tube, followed by systemic TXA, and group (C), the placebo group, received sodium chloride 0.9% with the same intravenous (IV) or nasogastric tube. Conventional IV fluid was administered to all patients and a conventional proton pump inhibitor (PPI) was initiated for patients with a probable clinical diagnosis of upper GIB. All participants and study personnel were blinded to the study arms. In this double-blind experiment, neither the patients who were administered drugs nor the researchers (injectors of the drugs) knew which participants belonged to the control (placebo) group, nor the test group. In fact, the content of the used agents was uncertain by coding the boxes and syringes. Only after all data were recorded, did the researchers learn which participants were which. The study protocol was approved by the ethics committee of the Iran University of Medical Sciences (IUMS number: 93-03-182-25168-106451), and the trial was also registered in Iranian Registry of Clinical Trials which was primary registry in the World Health Organization registry network (IRCT2014120120178N1).

Participants

For the 410 patients who enrolled in this study, GIB (hematochezia, melena, and hematemesis) was documented by history taking, physical examination, and/or lab tests. The exclusion criteria were: age <18 years, pregnancy, lactation, oral contraceptive pill use, history of a recent (12 months) thromboembolic disease, malignancy (except those cured and with no recurrence for >2 years), end-stage renal disease and/or nephrotic syndrome, underlying cardiac arrhythmia where anticoagulation is indicated (atrial fibrillation/flutter), oral body temperature >38℃, acquired color and/or acuity vision impairment, seizure, severe liver disease and/or previous history of variceal bleeding, unobtainable informed consent, and any allergic history to the medication.

Data collection

Demographic data, such as age, sex, smoking, type of GIB, vital signs, and laboratory data (hemoglobin (Hb), creatinine, prothrombin time, partial thromboplastin time, international normalized ratio, blood group/Rh), were recorded. The time to need for doing the first (urgent and/or early) endoscopy after admission (based on the expert's viewpoint and the patients' hemodynamic stability status) and endoscopic data, including endoscopy time, clearance, and findings, was recorded. The need for emergency endoscopy referred to performing this procedure in a time span of 24 h after presentation especially due to acute severe GI bleeding or any evidence of other problems such as acute biliary pancreatitis and acute cholangitis. The re-bleeding rate, number and type of transfusion, endoscopic and/or surgical intervention rate, ICU admission rate due to this condition, duration of ward/ICU admission, drug adverse effects, mortality rate, and causes of death were recorded in the first 24–72 h, and if possible, the patients were followed up for 1 month. The 1-month mortality rate and causes of death were recorded by calling the caregivers of the patients and/or hospitalization documents. Hemodynamic instability (shock, orthostatic hypotension) or active bleeding (manifested by hematemesis, bright red blood per nasogastric tube, or hematochezia) were indications for ICU admission and resuscitation. Urgent endoscopy (defined as first endoscopy <6 hours after an emergency visit) was performed after resuscitation.

Need to blood transfusions were defined as Hb < 7 g/dl (70 g/l) for most patients, with a goal of maintaining it at a level ≥7 g/dl (70 g/l). The threshold might be higher at Hb ≥ 9 g/dl (90 g/l) for high-risk patients with adverse events in the setting of considerable anemia.^17–19 However, need to blood transfusions were defined as the apparently normal Hb levels in the patients with active bleeding and hypovolemia.¹⁷

Re-bleeding was defined as overt hematemesis; passage of fresh blood from the rectum; a drop in Hb level of >2 g/l within any 24 h period after the first 24 h following endoscopic homoeostasis; shock (defined as a systolic blood pressure of ≤90 mmHg or a heart rate ≥110 beats/min) in the presence of continuing melena; or the presence of bright red blood in the stomach or duodenum, or both, at repeat endoscopy when more bleeding was suspected.¹⁸

Statistical analysis

Results were presented as mean ± standard deviation for quantitative variables and were summarized by frequency (percentage) for categorical variables. Continuous variables were compared using an analysis of variance or a Kruskal–Wallis H test whenever the data did not appear to have normal distribution or when the assumption of equal variances was violated across the study groups. Categorical variables were compared using the Chi-square test. All statistical analyses were conducted using the statistical software SPSS version 16.0 for windows (SPSS Inc., Chicago, IL, USA). A p-value <0.05 was considered statistically significant.

Results

In total, 410 patients were enrolled in the study. The mean age of the participants was 58.81 ± 15.60 years and 66.8% were male; age and sex distribution between groups was not statistically different. A total of 23.9% of our patients were smokers, with no significant differences in the three groups (p = 0.784). Hematemesis was the most common clinical presentation of GIB overall. No significant differences were seen in the three groups according to the symptom presentations (p = 0.185). Table 1 shows the clinical and demographic data of the patients in this trial. No differences were noted for the clinical data of the three groups (Table 1).

Table 1.

Baseline characteristics and clinical data of study population.

Item	A (n = 138)	B (n = 133)	C (n = 139)	Total (N = 410)	P-value
Age	60.50 ± 15.35	56.73 ± 20.42	59.14 ± 15.70	58.81 ± 15.60	0.241
Sex (Male)	85 (61.6)	97 (72.9)	92 (66.2)	274 (66.8)	0.138
Smoking	34 (24.6)	29 (21.8)	35 (25.2)	98 (23.9)	0.784
Hemoglobin (admission)	10.06 ± 2.97	9.88 ± 2.96	9.70 ± 2.73	9.86 ± 2.88	0.668
Creatinine(admission)	1.27 ± 0.47	1.32 ± 0.56	1.42 ± 0.69	1.34 ± 0.58	0.090
Systolic blood pressure (admission)	115.62 ± 21.84	115.90 ± 18.53	114.32 ± 16.72	115.27 ± 19.11	0.768
Pulse rate (admission)	90.97 ± 15.87	86.86 ± 15.61	87.37 ± 14.15	88.42 ± 15.29	0.054
GI Bleeding type					0.185
Hematemesis	45 (32.6)	29 (21.8)	41 (29.5)	115 (28.0)
Hematemesis and hematochezia	1 (0.7)	0 (0.0)	0 (0.0)	1 (0.2)
Hematemesis and melena	38 (27.5)	36 (27.1)	31 (22.3)	105 (25.6)
Hematochezia	12 (8.7)	25 (18.8)	18 (12.9)	55 (13.4)
Hematochezia and melena	2 (1.4)	2 (1.5)	0 (0.0)	4 (1.0)
Melena	40 (29.0)	41 (30.8)	49 (35.3)	130 (31.7)
Need to Transfusion	24 (17.5)	23 (17.3)	16 (11.5)	63 (15.4)	0.294
Mean of Unit Transfusion (admission)	1.3 ± 1.78	1.30 ± 2.35	1.01 ± 1.62	1.11 ± 1.94	0.385
Firstly admitted in ICU	2 (1.4)	7 (5.3)	8 (5.8)	17 (4.1)	0.380

ICU: intensive care unit; GI: gastrointestinal

Duodenal ulcer was the most common endoscopic finding overall (30.7%) and its incidence was not significantly different among the groups (Figure 1). The most common finding of colonoscopy overall was diverticular bleeding (4.9%), and its incidence did not differ significantly among the groups (Figure 2). We found esophageal varices for the first time in 11.9% (7/59) of the patients with GIB who underwent upper GI endoscopy.

Figure 1.

Endoscopic findings of intravenous tranexamic acid treatment group (A), intravenous plus topical tranexamic acid treatment group (B) and placebo group (C).

Figure 2.

Colonoscopic findings of intravenous tranexamic acid treatment group (A), intravenous plus topical tranexamic acid treatment group (B) and placebo group (C).

A total of 57 endoscopy interventions were done, including 13 adrenaline injections, 14 adrenaline injections and hemoclips, 23 adrenaline injections and argon plasma coagulation; and 7 esophageal variceal band ligations.

Comparison of the time to the first endoscopy differed significantly among the three groups (p < 0.001). The mean time to the first endoscopy was 17.70 ± 7.99 h overall, and was 18.35 ± 8.23, 19.28 ± 7.29, and 15.5 8 ± 7.99 hours in groups (A), (B), and (C) respectively (Table 2). A total of 20 patients in group (A) (14.49%) and 14 patients in group (B) (10.52%) required urgent endoscopy, compared with 42 patients in group (C) (30.21%) (p < 0.001).

Table 2.

Comparison of endoscopic data and time between three groups.

Item	A (n = 138)	B (n = 133)	C (n = 139)	Total (N = 410)	p-value
Endoscopy time	6.29 ± 3.04	6.69 ± 3.64	6.66 ± 2.39	6.55 ± 3.05	0.483
Endoscopy intervention	23/133 (17.3%)	16/130 (12.3%)	18/138 (13.0%)	57/401 (14.2%)	0.608
Time to endoscopy room (upper)	17.83 ± 8.41	18.94 ± 7.56	15.08 ± 7.83	17.70 ± 7.99	<0.001*
Time to endoscopy room (lower)	22.71 ± 4.68	20.56 ± 6.06	18.89 ± 8.49	17.70 ± 7.99	0.279
Urgent endoscopy rate (upper)	16/119 (13.4 %)	12/103 (11.7 %)	41/120 (34.2%)	76/401 (18.9%)	<0.001*
Urgent endoscopy rate (lower)	0/14 (0%)	2/27 (7.4%)	5/18 (27.8%)	76/401 (18.9%)	0.34
Difficulty in endoscopy					0.238
Active bleeding	5 (3.7%)	3 (2.3%)	6 (4.3%)	14 (3.4%)
Blood oozing	3 (2.2%)	3 (2.3%)	12 (8.6%)	18 (4.4%)
Clot	11 (8.2%)	12 (9.2%)	18 (13.0%)	41 (9.9%)
Clear	115 (85.8%)	112 (86.1%)	102 (73.9%)	329 (80.2%)
Not done	5 (3.6%)	3 (2.3%)	1 (0.7%)	9 (2.0%)

Mild non-serious adverse effects were seen in 14.1% of all patients. Nausea (7.8%) was the most common side effect. At 3 days after treatment, the mild drug side effects were higher in group (A) than in group (B). At 1 month after the medical intervention, the thromboembolic events were significantly more prevalent in control group (C) (5/138) than in group (A) (1/137) (p = 0.032). No thromboembolic events were observed in group (B).

The re-bleeding, blood transfusion, and endoscopic and/or surgical intervention rates were not statistically different among the groups (Table 3).

Table 3.

Clinical outcomes at 72 hours and 1-month follow up.

Item	A (n = 138)	B (n = 133)	C (n = 139)	Total (N = 410)	p-value
72 hours
Mortality rate	1/138 (0.7%)	2/133 (1.5%)	6/139 (4.3%)	9/410 (2.1%)	0.886
Re-bleeding rate	9/138 (6.5%)	11/133 (8.3%)	13/139 (9.4%)	33/410 (8.1%)	0.682
Re-bleeding time (h)	43.78 ± 24.34	58.91 ± 18.93	49.23 ± 20.48	50.97 ± 21.33	0.276
Surgical intervention	4/138 (2.9%)	3/132 (2.3%)	2/139 (1.4%)	9/409 (2.2%)	0.708
Drug side effects	31/138 (22.5%)	24/133 (18.0%)	3/139 (2.2%)	58/410 (14.1%)	<0.001*
Ward admission duration (d)	3.50 ± 1.21	4.01 ± 1.66	3.86 ± 1.92	3.79 ± 1.63	0.034*
ICU admission (d)	3.30 ± 1.42	4.33 ± 5.19	4.67 ± 6.45	4.23 ± 5.15	0.802
ICU admission rate	11/138 (8.0%)	15/131 (11.5%)	18/139 (12.9%)	44/408 (10.8%)	0.392
ICU admission due to bleeding	8 (5.8%)	8 (6.0%)	10 (7.2%)	26 (6.3%)	0.380
1 month
Mortality rate	9/115 (7.8%)	11/115 (9.6%)	18/119 (15.1%)	38/349 (10.9%)	0.172
Re-bleeding rate	2/133 (1.5%)	1/131 (0.8%)	1/138 (0.7%)	4/349 (1.1)	0.779
Surgical intervention	4/138 (2.9%)	4/132 (3.0%)	2/138 (1.4%)	10/407 (2.5%)	0.641
Thromboembolic events	1/137 (0.7%)	0/132 (0.0%)	5/138 (3.6%)	6/349 (1.7%)	0.032*
Need to transfusion	2/133 (1.5%)	1/131 (0.8%)	1/138 (0.7%)	4/349 (1.1)	0.779

statistically significant difference

ICU: intensive care unit

The overall and based on the bleeding, 1-month mortality rate were 10.9% and 2.2% respectively. The prevalence of bleeding, as a cause of mortality in patients with GIB, was lower in group (B) than in groups (A) and (C) (Figure 3). The other causes of mortality showed no significant differences among the three groups.

Figure 3.

Mortality causes in intravenous tranexamic acid treatment group (A), intravenous plus topical tranexamic acid treatment group (B) and placebo group (C).

Discussion

GIB requires emergency care that includes medical treatment and endoscopy to find the source of bleeding and treat it by endoscopic intervention, if possible.¹⁹ Gastroenterology practices must therefore be available 24 hours a day to manage GI emergencies like GIB, which is seen in most cases referred to the emergency room.²⁰ Airway management and hydration with intravenous fluids or blood transfusion are the primary goals of GIB management.²¹ Based on the bleeding source, different medications may be associated with better outcomes.²¹

Oral and intravenous PPIs have been considered as a cost-effective medication in upper GIB management to reduce re-bleeding, mortality, and the need for surgery.^22,23 Radiologic and endoscopic intervention methods were done for diagnosis and treatment of patients with GIB.

Some experimental studies have suggested that TXA, a clot breakdown inhibitor, is effective for GIB management.¹² TXA inhibits plasminogen, which is a main enzyme involved in fibrinolysis. TXA also directly reduces the fibrinolytic activity of pepsin. Recently, TXA has also been shown to have anti-inflammatory effects.²⁴ TXA has been administered in upper GIB patients and the highest plasma concentration was seen 1 h after injection.¹² The most common GI adverse side effects associated with TXA administration are diarrhea and nausea. Dizziness and hypotension can be seen following rapid injection of TXA, when the recommended infusion rate is <100 mg/min.

The antifibrinolytic effects of TXA have led to its use as an effective drug for reducing bleeding following surgery, heavy menstruation, and even other types of serious post-traumatic bleeding. The recommended dosage for TXA is 0.5–1 g (10–15 mg/kg), delivered intravenously 3–4 times per day.²⁵ In this way, we found an overall incidence of mild, non-deleterious adverse effects of 14.1%. Similar to previous studies,²⁶ the most common adverse effects were GI problems (nausea 7.8%), as a common presentation in GI bleeding patients. We recorded an incidence of 2.2% for GI problems in the control group. It seems that an exact estimation of side effects is not possible in this bleeding group. Nevertheless, these problems may not be considered as side effects of TXA, because all our patients had underlying GI problems.

Thromboembolic events are rare, serious and life-threatening adverse effect of TXA. The CRASH-2 trial demonstrated a reduced risk of death, without any thromboembolic adverse events, in 20,211 trauma patients who received TXA.¹¹ In the present study, no thromboembolic event was documented during the first week after TXA administration. But we recorded 1 and 5 thromboembolic events in group (A) and (C) patients, respectively after 1 month of follow up. In comparison with our results, some case reports have shown thromboembolic events associated with TXA.^27,28 Thus, these findings should be confirmed by other surveys with larger populations.

The therapeutic and diagnostic role of endoscopy remains critical in GIB management unless the condition requires emergent intervention.

Contrary to our expectation, urgent endoscopy was less effective than early endoscopy in improving outcome. Some risks of urgent upper endoscopy are aspiration, adverse reactions to conscious sedation, unclear field due to active bleeding, risk of perforation and increasing bleeding while attempting therapeutic intervention and the difficulty in achieving homeostasis.²⁹ Urgent endoscopy, when compared with elective endoscopy, has more side effects, lower accuracy and availability, and leads to more misdiagnoses. Hence, a consensus exists regarding the need to lower urgent endoscopy in patients with GIB. And rapid endoscopy should be done on a case-by-case basis.⁷

The days of admission and transfusion rates were lowered by an early endoscopic approach in previous studies.³⁰ The efficacy of treatment and risk stratification by early endoscopy remains controversial with respect to the reduction of resource utilization or negative patient outcomes.^31–33 Some previous studies have shown improved outcomes and lowered resource utilization following early endoscopy^33,34 while other studies did not.^31,32

Endoscopy cannot reveal the cause of upper GIB in about 20% of patients.³⁵ We found normal endoscopies in 10.2% patients (40/390) in this study.

Early administration of TXA can apparently decrease the urgent endoscopy rate and increase the time for doing the first procedure in acute significant GIB patients. This means that its antifibrinolytic effect can help the physicians and patients in a decision to change urgent to elective procedures, with better outcomes for both physicians and patients. According to endoscopic findings, the occurrence of active bleeding, oozing, and clotting was slightly higher in the placebo group. For this reason, a previous study has suggested the use of TXA in GIB,³⁶ although its efficacy in lowering urgent endoscopy rate was not investigated. In the present study, re-bleeding, endoscopic and surgical intervention, and blood transfusion rates were not significantly lowered, consistent with the meta-analysis conducted by Bennett et al.¹² According to our study, the mortality rates from all causes and based on bleeding were slightly higher in the placebo group (C), but the differences were not statistically significant among the treatment groups, in contrast with this earlier meta-analysis.¹² However, the results of the meta-analysis were not reliable due to high dropout rates in some trials, and the authors suggested that more studies should be done focusing on the efficacy of TXA for the treatment of GIB.

Based on the subgroup analysis in each upper and lower GIB patients (Table 2), it seems that the topical TXA effect can slightly decrease urgent endoscopy rates in upper against lower GIB patients, as we expected.

This study has some limitations: First, 61 out of 410 patients failed to continue their 1-month follow up. Second, we recorded 1-month mortality rates and causes of death by calling the caregivers of the patients and/or reviewing hospitalization documents. Therefore, we could not precisely evaluate the cause of death. However, the mortality rate and the causes of death after early administration of TXA was not the main aim of this study.

In summary, our results show that the antifibrinolytic properties of TXA, the same as the effect of PPIs in clot stabilizing, can help physicians and patients in the decision to change an urgent endoscopy to an elective procedure, with better outcomes for both. The safety and low cost of TXA should encourage its use beyond the current standard treatment.

Further research with larger populations is needed to elucidate other aspects of management of GIB patients with TXA.

Footnotes

Declaration of conflicting interests

The authors declare no conflicts of interest in preparing this article.

Ethics approval

Written informed consent was obtained from all patients and study protocol was approved by ethics committee of Iran University of Medical Sciences, Tehran, Iran.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Informed consent

None.

References

Button

Roberts

Evans

et al.

Hospitalized incidence and case fatality for upper gastrointestinal bleeding from 1999 to 2007: A record linkage study. Aliment Pharmacol Ther 2011; 33: 64–76.

Fallah

Prakash

Edmundowicz

. Acute gastrointestinal bleeding. Med Clin North Am 2000; 84: 1183–1208.

Longstreth

. Epidemiology of hospitalization for acute upper gastrointestinal hemorrhage: A population-based study. Am J Gastroenterol 1995; 90: 206–210.

Srygley

Gerardo

Tran

et al.

Does this patient have a severe upper gastrointestinal bleed?

JAMA 2012; 307: 1072–1079.

Masoodi

Saberifiroozi

. Etiology and outcome of acute gastrointestinal bleeding in Iran: A review article. Middle East J Dig Dis 2012; 4: 193–193.

Laine

McQuaid

. Endoscopic therapy for bleeding ulcers: an evidence-based approach based on meta-analyses of randomized controlled trials. Clin Gastroenterol Hepatol 2009; 7: 33–47.

Targownik

Murthy

Keyvani

et al.

The role of rapid endoscopy for high-risk patients with acute nonvariceal upper gastrointestinal bleeding. Can J Gastroenterol 2007; 21: 425–425.

Brown

Birkmeyer

O’Connor

. Meta-analysis comparing the effectiveness and adverse outcomes of antifibrinolytic agents in cardiac surgery. Circulation 2007; 115: 2801–2813.

Leminen

Hurskainen

. Tranexamic acid for the treatment of heavy menstrual bleeding: efficacy and safety. Int J Wom Health 2012; 4: 413–413.

10.

Cheriyan

Maier

Bianco

et al.

Efficacy of tranexamic acid on surgical bleeding in spine surgery: a meta-analysis. Spine J 2015; 15: 752–761.

11.

Shakur H, Roberts I, et al. CRASH-2 trial collaborators. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010; 376: 23–32.

12.

Bennett

Klingenberg

Langholz

et al.

Tranexamic acid for upper gastrointestinal bleeding. Cochrane Database Syst Rev 2014; 11: CD006640–CD006640.

13.

Ker

Beecher

Roberts

. Topical application of tranexamic acid for the reduction of bleeding. Cochrane Libr 2013; 23: CD010562–CD010562.

14.

Rafeey

Shoaran

Ghergherechi

. Topical tranexamic acid as a novel treatment for bleeding peptic ulcer: A randomised controlled trial. Afr J Paediatr Surg 2016; 13: 9–9.

15.

Roberts

Coats

Edwards

et al.

HALT-IT-tranexamic acid for the treatment of gastrointestinal bleeding: study protocol for a randomised controlled trial. Trials 2014; 15: 450–450.

16.

Hawkey

Cole

McIntyre

et al.

Drug treatments in upper gastrointestinal bleeding: value of endoscopic findings as surrogate end points. Gut 2001; 49: 372–379.

17.

Laine

Jensen

. Management of patients with ulcer bleeding. Am J Gastroenterol 2012; 107: 345–360.

18.

Barkun

Sabbah

Enns

et al.

The Canadian Registry on Nonvariceal Upper Gastrointestinal Bleeding and Endoscopy (RUGBE): Endoscopic hemostasis and proton pump inhibition are associated with improved outcomes in a real-life setting. Am J Gastroenterol 2004; 99: 1238–1246.

19.

Hreinsson J.P, Kalaitzakis E, Gudmundsson S, et al. Upper gastrointestinal bleeding: incidence, etiology and outcomes in a population-based setting. Scand J Gastroenterol 2013; 48: 439–447.

20.

Sonnenberg

. Timing of endoscopy in gastrointestinal bleeding. United European Gastroenterol J 2014; 2: 5–9.

21.

Jairath

Barkun

. The overall approach to the management of upper gastrointestinal bleeding. Gastrointest Endosc Clin N Am 2011; 21: 657–670.

22.

Leontiadis G, Sreedharan A, Dorward S, et al. Systematic reviews of the clinical effectiveness and cost effectiveness of proton pump inhibitors in acute upper gastrointestinal bleeding. 2007; 11: iii–iv, 1–164.

23.

Tsoi

Hirai

Sung

. Meta-analysis: comparison of oral vs. intravenous proton pump inhibitors in patients with peptic ulcer bleeding. Aliment Pharmacol Ther 2013; 38: 721–728.

24.

Nutbeam

. In adult patients presenting as emergencies with upper gastrointestinal bleeding, does tranexamic acid decrease mortality? Afr J Emer Med 2015; 5: 85–92.

25.

Tengborn

. Fibrinolytic inhibitors in the management of bleeding disorders. Treatment of Hemophilia 2007; 42: 1–15.

26.

Sentürk

Cakmak

Yildiz

et al.

Tranexamic acid for cesarean section: a double-blind, placebo-controlled, randomized clinical trial. Arch Gynecol Obstet 2013; 287: 641–645.

27.

Rydin

Lundberg

. Tranexamic acid and intracranial thrombosis. Lancet 1976; 308: 49–49.

28.

Taparia

Cordingley

Leahy

. Pulmonary embolism associated with tranexamic acid in severe acquired haemophilia. Eur J Haematol 2002; 68: 307–309.

29.

Boţianu

Matei

Tanţău

et al.

Urgent versus early endoscopy in high risk patients with acute upper gastrointestinal bleeding: A comparative study in a tertiary center with a permanent endoscopy call. Rom J Intern Med 2013; 51: 35–40.

30.

El-Tawil

. Trends on gastrointestinal bleeding and mortality: Where are we standing? World J Gastroenterol 2012; 18: 1154–1154.

31.

Bjorkman

Zaman

Fennerty

et al.

Urgent vs. elective endoscopy for acute non-variceal upper-GI bleeding: an effectiveness study. Gastrointest Endosc 2004; 60: 1–8.

32.

Sarin

Monga

Adams

. Time to endoscopy and outcomes in upper gastrointestinal bleeding. Can J Gastroenterol 2009; 23: 489–489.

33.

Wysocki

Srivastav

Winstead

. A nationwide analysis of risk factors for mortality and time to endoscopy in upper gastrointestinal haemorrhage. Aliment Pharmacol Ther 2012; 36: 30–36.

34.

Tsoi

Chiu

Chan

et al.

The risk of peptic ulcer bleeding mortality in relation to hospital admission on holidays: A cohort study on 8222 cases of peptic ulcer bleeding. Am J Gastroenterol 2012; 107: 405–410.

35.

Palmer

Nairn

. Management of acute gastrointestinal blood loss: summary of SIGN guidelines. BMJ 2008; 10: 337–337.

36.

Sreedharan

Martin

Leontiadis

et al.

Proton pump inhibitor treatment initiated prior to endoscopic diagnosis in upper gastrointestinal bleeding. Cochrane Libr 2010; (7): CD005415–CD005415.